1. Field of the Invention
The present invention relates to a semiconductor device and a method of manufacturing a semiconductor device, and in particular, to a structure of a junction terminating region portion of a power semiconductor device which is suitable for a switching element for electric power.
2. Related Background Art
In response to the demand to make electrical power equipment compact and high-performance in recent years in the power electronics field, performance improvements with respect to making lowering loss, increasing speed and improving the ruggedness have been carried out in addition to making them have a high breakdown voltage and making them able to handle great current in the power semiconductor devices. Among them, a power MOSFET (Metal Oxide Semiconductor Field Effect Transistor) has been established as a key device in the switching power source field or the like because of the high-speed switching performance thereof.
Because the MOSFET is a majority carrier device, the MOSFET has an advantage that there is no minority carrier storage time and switching is fast. However, on the other hand, because there is no electric conductivity modulation, a device which has a high breakdown voltage is disadvantageous with respect to the ON-state resistance as compared with a bipolar device such as an IGBT (Insulated Gate Bipolar Transistor) or the like. This results from the fact that the higher the breakdown voltage a device has, the more the ON-state resistance of the MOSTFET increases, because it is necessary to make an n type base layer thicker and to make the impurity concentration lower in order to obtain a higher breakdown voltage in the MOSFET.
The ON-state resistance of the power MOSFET greatly depends on the electrical resistance in a conductive layer (n type drift layer) portion. Further, the impurity concentration determining the electrical resistance at the n type drift layer cannot rise to greater than or equal to the limit, in accordance with the breakdown voltage of the pn junction which the p type base and the n type drift layer form. Therefore, a trade-off relationship exists between the device breakdown voltage and the ON-state resistance. It is important to improve this trade-off for a low electric power consumption device. In this trade-off, there is a limit which is determined by the material of the device, and this limit must be exceeded in order to realize a low ON-state resistance device exceeding existing power devices.
As one example of MOSFETs for solving this problem, a structure is known in which a resurf structure called a super junction structure is buried in an n type drift layer. A conventional power MOSFET having a super junction structure will be described with reference to FIG. 36. Note that, in the following figures, like parts are denoted by like reference numerals, and detailed descriptions thereof will be omitted.
FIG. 36 is a cross-sectional view showing a schematic structure of one example of a conventional power MOSFET. In the MOSFET shown in the view, an n+ type drain layer 100 is formed on one surface of an n− type drift layer 102, and a drain electrode 40 is formed on the n+ type drain layer 100. Further, a plurality of p type base layers 108 are selectively formed in the other surface portion of the n− type drift layer 102, and n+ type source layers 110 are selectively formed in the surfaces of the respective p type base layers 108. Further, a gate electrode 114 is formed on the surface region of the n− type drift layer 102 which are sandwiched by the adjacent p type base layers 108, the surfaces of the p type base layers 108 sandwiching the n− type drift layer 102, and the surface region of the portions of the n+ type source layer 110 facing each other via the p type base layers 108 and the n− type drift layer 102, with a gate insulating film 112 interposed therebetween. Further, source electrodes 116 are respectively formed on the region of the surfaces of the n+ type source layer 110 and the surface of the p type base layer 108 so as to sandwich the gate electrode 114. Moreover, in the n− type drift layer 102 between the p type base layer 108 and the n+ type drain layer 100, a p type drift layer 106, which is formed so as to form a resurf layer and is connected to the p type base layer 108, is formed. In this way, the power MOSFET shown in FIG. 36 has a vertical type resurf structure in which the p type drift layers 106 and the portions of the n− type drift layers 102 sandwiched by these p type drift layers 106 are alternately repeated in the lateral direction.
In an OFF-state, a depletion layer spreads at junctions between these p type drift layers 106 and n− type drift layers 102. Even if the impurity concentration of the n− type drift layers 102 is made high, the n− type drift layers 102 and the p type drift layers 106 are completely depleted before breaking down. In accordance therewith, a breakdown voltage which is the same as that of the conventional MOSFET can be obtained.
The impurity concentration of the n− type drift layer 102 does not depend on a breakdown voltage of the device, but it depends on the width of the p type drift layer 106 and the width of the n− type drift layer 102 itself between these p type drift layers 106. If the width of the n− type drift layer 102 and the width of the p type drift layer 106 are made narrower, the impurity concentration of the n− type drift layer 102 can be made much higher, and a greater reduction of the ON-state resistance and a higher breakdown voltage can be achieved.
At the time of designing such a MOSFET, the impurity concentrations of the n− type drift layer 102 and the p type drift layer 106 are important to determine the breakdown voltage and the ON-state resistance. In principle, due to the respective impurity concentrations of the n− type drift layer 102 and the p type resurf layer 106 being made equal, the impurity concentrations equivalently become zero, the high breakdown voltage can be obtained.
However, with respect to the semiconductor device having a conventional super junction structure, no structure has been developed which is effective for obtaining a high breakdown voltage in a blocking state (OFF-state) or at the time of turning-off at a junction terminating region portion which is positioned at the outer periphery of an element active region portion (hereinafter referred to as a cell region portion) and which is a region portion for maintaining the breakdown voltage by attenuating the electrical field by extending the depletion layers. Therefore, because the way of spreading of the depletion layers in the cell region portion and the junction terminating region portion are different from one another, the optimum impurity concentrations are different from one another. Accordingly, if the device is manufactured such that the impurity amounts in the cell region portion and the junction terminating region portion are the same, the breakdown voltage at the terminating portion decreases, and an electric field locally concentrates at this place. As a result, there are cases in which the device is broken. In this way, there is the problem that a sufficiently high breakdown voltage cannot be obtained by the entire device in the prior art.
Further, because there are dispersions among the processes at the time of actual manufacturing, it is difficult to make the respective impurity amounts of the n− type drift layer 102 and the p type drift layer 106 completely equal, and the breakdown voltage deteriorates in accordance therewith. Accordingly, it is necessary to carry out designing of the device in consideration of such a decrease of the breakdown voltage due to the process margin. In order to lower the ON-state resistance, it is effective to raise the impurity concentration of the n− type drift layer 102. On the other hand, the process margin for the breakdown voltage is determined by the difference in the impurity amounts between the n− type drift layer 102 and the p type drift layer 106. Therefore, when the impurity amount of the n− type drift layer 102 is increased, because the difference itself determining the process margin is not changed, the ratio between the allowed impurity amount and the impurity amount of the n− type drift layer 102 becomes small. Namely, the process margin becomes small.